Leveraging Query Executions to Improve Index Recommendations

1. A system for selecting an execution cost efficient query plan, comprising: at least one processor; and memory including instructions that, when executed by the at least one processor, cause the at least one processor to: receive a first query plan and a second query plan for performing a query with a database, wherein the first query plan is different from the second query plan; use a classifier model to determine relative execution cost efficiency between the first query plan and the second query plan, wherein the classifier model is trained using pairs of query plans and an indicator of which query plan of each pair of query plans has a more efficient relative execution cost for the database; and output a ranking of the first query plan and second query plan, wherein the first query plan and second query plan are ranked based on execution cost efficiency.

2. The system of claim 1 , wherein the first query plan is based on a first index configuration for the database and the second query plan is based on a second index configuration for the database.

3. The system of claim 1 , wherein a set of features are identified for each pair of query plans.

4. The system of claim 3 , wherein a weight is applied to each feature of the set of features.

5. The system of claim 4 , wherein the set of pairs of query plans are converted into feature vectors using the respective set of features of each pair of query plans.

6. The system of claim 1 , wherein the output includes a certainty value indicating a confidence level for the ranking.

7. The system of claim 1 , wherein the classifier model is trained using relative execution cost comparisons between a set of pairs of query plans of a second database.

8. The system of claim 7 , wherein the set of pairs of query plans of the database is less than the set of pairs of query plans of the second database.

9. The system of claim 8 , wherein training the classifier model is based on the set of pairs of query plans of the database and the set of pairs of query plans of the second database.

10. The system of claim 1 , further comprising instructions to: determine the first query plan is ranked as more cost efficient than the second query plan; select the first query plan; configure the database based on the first query plan; and execute a query based on the first query plan.

11. A method for selecting an execution cost efficient query plan, comprising: receiving a first query plan and a second query plan for performing a query with a database, wherein the first query plan is different from the second query plan; using a classifier model to determine relative execution cost efficiency between the first query plan and the second query plan, wherein the classifier model is trained using pairs of query plans and an indicator of which query plan of each pair of query plans has a more efficient relative execution cost for the database; and outputting a ranking of the first query plan and second query plan, wherein the first query plan and second query plan are ranked based on execution cost efficiency.

12. The method of claim 11 , wherein the first query plan is based on a first index configuration for the database and the second query plan is based on a second index configuration for the database.

13. The method of claim 11 , wherein the classifier model is trained using relative execution cost comparisons between a set of pairs of query plans of a second database.

14. The method of claim 13 , wherein the set of pairs of query plans of the database is less than the set of pairs of query plans of the second database.

15. The method of claim 14 , wherein training the classifier model is based on the set of pairs of query plans of the database and the set of pairs of query plans of the second database.

16. At least one non-transitory computer readable medium including instructions for selecting an execution cost efficient query plan that when executed by at least one processor, cause the at least one processor to: receive a first query plan and a second query plan for performing a query with a database, wherein the first query plan is different from the second query plan; use a classifier model to determine relative execution cost efficiency between the first query plan and the second query plan, wherein the classifier model is trained using pairs of query plans and an indicator of which query plan of each pair of query plans has a more efficient relative execution cost for the database; and output a ranking of the first query plan and second query plan, wherein the first query plan and second query plan are ranked based on execution cost efficiency.

17. The at least one computer readable medium of claim 16 , wherein the first query plan is based on a first index configuration for the database and the second query plan is based on a second index configuration for the database.

18. The at least one computer readable medium of claim 16 , wherein the classifier model is trained using relative execution cost comparisons between a set of pairs of query plans of a second database.

19. The at least one computer readable medium of claim 18 , wherein the set of pairs of query plans of the database is less than the set of pairs of query plans of the second database.

20. The at least one computer readable medium of claim 19 , wherein training the classifier model is based on the set of pairs of query plans of the database and the set of pairs of query plans of the second database.